Assessment of the Female ASD Phenotype: The M-ASD Questionnaire
ASD diagnostics are based mainly on the male ASD phenotype. People with a female ASD phenotype are often under- or misdiagnosed. This phenotype seems more subtle due to better camouflaging techniques (Lai et al., 2011), a higher social motivation and gender specific preoccupations (Hiller et al., 2014). However, an ASD diagnosis is indisputably present.
A questionnaire taking into account the female ASD phenotype could aid in a faster identification of these women, but also of men with this more subtle phenotype. This could lead to a better prognosis, prevent secondary problems, reduce family stress and societal costs (García-Primo et al., 2014).
In 2016, we developed the Miss-ASD questionnaire (M-ASD). It consists of 120 items derived from literature search on female ASD expressions, clinical impressions of the authors, and data analysis of sex differences in adults with ASD on other questionnaires. The M-ASD covers 6 domains: Social interaction and communication, Rigidity, Coping and camouflaging, Sensory issues, Information processing, and Miscellaneous (INSAR, Grondhuis et al., 2018).
Shortening the M-ASD for clinical use.
Adults suspected for ASD underwent an extensive ASD diagnostic assessment. At start, they also completed the M-ASD. Independent researchers scored these questionnaires.
The research group ultimately consisted of 183 patients (age: M=35.07, SD=12.54), of which ultimately 83% received an ASD diagnosis. There were 88 women with ASD, 63 men with ASD, and also 25 women and 7 men with other psychiatric diagnoses (non-ASD group).
Qualitative analyses were based on recent literature (2016-2018) and two focus groups: 1) 3 women and 1 men with ASD, and 2) 4 psychologists with elaborate ASD expertise. Quantitative analyses consisted of individual item analyses, including discriminant indices and Fisher-Z exact tests. Item judgement, based on focus groups and statistical analyses, was categorized as a ‘keep’, ‘reject’ or ‘doubt’ classification.
ASD patients scored higher than non-ASD patients on every M-ASD item. Based on the combination of focus groups and statistics, the best 47 items were retained. Among them are the items that differentiated the most clearly between ASD/non-ASD, and more specifically between ASD/non-ASD women (DI 0.3-0.7, significant Fisher test). The items that best captured the female ASD phenotype (differentiation ASD women/ ASD men) concern sensory and camouflage issues.
Based on the focus groups and new literature, 15 items of the 47 items were rephrased and 7 novel items, mostly on camouflaging, were added.
The new, abbreviated version of the M-ASD (54 items) is more appropriate for clinical use than the initial version. The M-ASD seems to be able to capture the female ASD phenotype. The most prominent items concern sensory and camouflaging issues, as was expected from literature.
Most ASD screening instruments are developed and validated in the general population. Those instruments perform less well in clinical practice, in terms of sensitivity and specificity (Bezemer et al., submitted). This is not the case for the M-ASD. Follow-up research will focus on the further validation and standardization of this promising new screening tool in a larger clinical sample, also including non-ASD samples.